基于IPPG技术的情绪识别算法

Emotion Recognition Algorithm Based on IPPG Technology

基于成像式光电容积描记技术(Imaging Photoplethysmography,IPPG)的情绪识别具有非接触式和客观性等优势。但基于IPPG信号的情绪识别还处于探索阶段,因IPPG信号质量容易受到影响,识别情绪的准确率还有待提高。为解决上述问题,选择额头作为感兴趣区域,通过小波变换和带通滤波提高信号质量,提取心率变异性特征,并使用支持向量机、K近邻、决策树和随机森林对特征进行分类。在DEAP数据库上进行实验,结果表明:支持向量机的情绪识别效果最好,在唤醒维度上的准确率为61.09%,在效价维度上的准确率为53.31%。该研究在远程医疗、人机交互等领域有广阔的应用前景。

Imaging photoplethysmography(IPPG)technology has the advantages of non-contact and objectivity in monitor-ing physiological parameters.Emotion recognition based on IPPG signals is still in the exploration stage,and the quality of IPPG signals is easily affected,and the accuracy of recognizing emotions still needs to be improved.To solve the above problems,the forehead is selected as the region of interest,the signal quality is improved by wavelet transform and band-pass filtering,the heart rate variability features are extracted,and the features are classified using support vector machines,K-nearest neighbors,decision trees,and random forests.Experiments were conducted on the DEAP database,and the results showed that the support vector machine was the best for emotion recognition,with an accuracy of 61.09%on the arousal dimension and 53.31%on the valence dimension.The research has broad application prospects in the fields of telemedicine and human-computer interaction.